An obligate step in the life cycle of HIV is the integration of the viral DNA into the host chromosome. This reaction is mediated by the integrase enzyme (In) through the formation of a staggered double strand break (DSB) in the host DNA and covalent linkage of the viral DNA ends to the host DNA. We are developing yeast and E. coli based genetic detection systems to investigate in vivo mechanisms of In, as well as means of altering the reaction. We have examined the effects of overexpression of HIV in wild-type and various mutant yeast strains and have found no change in growth, morphology, viability, or sensitivity to DNA damaging agents. To increase the sensitivity of our assays we constructed a targeted In fusion protein. This consists of the yeast GAL4 DNA binding domain fused to the 5' end of the In protein. Work by others has shown i)fusions to this end of In are still active for in vivo assays, and ii) that GAL4 DNA binding domain fusion proteins are nuclear localized. We are testing this construct not only in two-hybrid screens, but also in the genetic assays described above. We are also preparing a plasmid with 3 repeats of the target sequence to examine plasmid loss. These repeats will also be placed next to one of two ura3 heteroalleles to determine if cutting by In stimulates recombination. These assays will be used to screen for human cDNAs, drugs, and yeast mutants that alter the In reaction in yeast. In E. coli we have found that low level over-expression of a soluble form of In is lethal. Lethality is not preceded by SOS induction, and overexpression of In mutants is also lethal. We are currently co-expressing In and Ini1 (which binds to In) in cells to determine if the lethality of In is modified by Ini1 and examining deletion mutants of In to determine which region causes killing.